I us ] 28 



About fifty of these larvne, for example, wliicli I reared last fall, and 

 all of which entered the gTound at about the same time, or in the course 

 of four days, in the first week of August, began to emerge in the winged 

 form on the tenth of the following June, and continued to come out till 

 the end of the first week of July, and I have known the apple variety 

 to emerge as late as the 18th of July. As these insects are of so large 

 and conspicuous a size, and as the larvse subsist upon the foliage of so 

 great a variety of trees, they furnish one of the best examples in the 

 whole class of insects, of phytophagic variation, or the variation of 

 color, both of the larvfe and of the perfect insects, according to the di- 

 versities of their food -plants. As might be expected, some of these va- 

 rieties differ but slightly from each other, whilst others, and especially 

 those which feed upon the black walnut, tliffer so remarkably from the 

 rest that they might almost be regarded as distinct species. Indeed, 

 some recent writers upon our Lepidoptera have elevated most of these 

 varieties to the rank of species, and have accordingly given to them dis- 

 tinct specific names. 



In speaking of the two allied moths, the Phycita nebiilo and the 

 Fhycita juglandis, in my second report, I had occasion to allude to the 

 vexed question in natural history, whether there be any essential distinc- 

 tion between what are known as species and varieties ; or whether these 

 terms only exj)ress different degrees of permanency in allied organic forms, 

 all of which are subject to change. Whatever opinion we may enter- 

 tain upon this difficult subject, the practical question will often arise, in 

 descriptive entomology, whether we shall regard and describe a particu- 

 lar kind of insect as a new species with a distinctive specific name, or 

 let it stand as only a variety of another and previously known species. 

 The well Imown test, applied to the higher aninuils, is that of uniting 

 the sexes of the two allied forms, and observing whether the progeny 

 have, or have not, the power of procreating their kind ; tlie former 

 being understood to indicate only varieties, whilst if the progeny are 

 sterile, it is supposed to prove that the parents were of diflerent sjK'cies. 

 But the application of this test requires so much time and attention 

 that it can scarcely be expected to be practiced merely for the purpose 

 of determining the specific identity of two allied insects. 



A criterion often applicable to insects is that founded uj)on their food- 

 habits, in accordance with the common rule that though the same species 

 of insect wiU often feed upon closely allied plants, they will very rarely 

 subsist upon plants of different natural families. But there are too 

 many well known exceptions to this rule to admit of making it a test of 

 specific identity. Indeed the larvie of many of the moths belonging to 

 the family of Arctiidse, pay no respect to this distinction ; and some of 

 them, of which the woolly hear or larva of the Sjnlosoma virginica is a 



